The present invention relates to a batch, semi-batch, continuous or semi-continuous process for converting a multihydroxylated-aliphatic hydrocarbon or an ester thereof to a chlorohydrin. More specifically, the present invention relates to a process wherein the multihydroxylated-aliphatic hydrocarbon or an ester thereof is a glycerol; and the chlorohydrin is a dichlorohydrin and an ester thereof, for example, 1,3-dichloro-2-propanol and/or 2,3-dichloro-1-propanol (DCH isomers). The process of the present invention has a benefit of producing chlorohydrins having a low concentration of volatile, by-product of halogenated hydrocarbon including chloroacetone. Chlorohydrins prepared by the process of the present invention, are useful in preparing epoxides such as epichlorohydrins. Unexpectedly, compositions of chlorohydrins and dichlorohydrins of glycerol have been found to have combined concentrations of volatile chlorinated hydrocarbon by-products and chloroacetone less than 2000 ppm throughout any stage of the said process.
Epichlorohydrin is a widely used precursor to epoxy resins. Epichlorohydrin is a monomer which is commonly used for the alkylation of para-bisphenol A; the resultant diepoxide, either as a free monomer or oligomeric diepoxide, may be advanced to high molecular weight resins which are used for example in electrical laminates, can coatings, automotive topcoats and clearcoats.
A known process for the manufacture of epichlorohydrin involves hypochlorination of allyl chloride to form dichlorohydrin. Ring closure of the dichlorohydrin mixture with caustic affords epichlorohydrin which is distilled to high purity (>99.6%). This chlorohydrin process requires two equivalents of chlorine and one equivalent of caustic per molecule of epichlorohydrin.
In another known process for producing epichlorohydrin the first step involves installing oxygen in the allylic position of propylene, via a palladium catalyzed reaction of molecular oxygen in acetic acid. The resulting allyl acetate is then hydrolyzed, chlorinated and the incipient dichlorohydrin is ring closed with caustic to epichlorohydrin. This process avoids the production of allyl chloride and therefore uses less chlorine (only one equivalent).
Both known processes for the manufacture of epichlorohydrin described above require the sacrificial use of chlorine, and complications associated with the industrial use and generation of hypochlorous acid (HOCl) can be magnified at industrial scale and these processes are known to produce substantial amounts of chlorinated by-products. In particular, it is well known that the hypochlorination of allyl chloride produces 1,2,3-trichloropropane and other undesirable chlorinated ethers and oligomers (RCls). RCl issues are managed as an increased cost to manufacture. As new capital is added to accommodate greater global production, a substantial investment in downstream processing must be added to accommodate and remediate these unwanted by-products. These same problems are analogous in the HOCl routes to propylene and ethylene chlorohydrin, and thus, these routes are less practiced.
An alternative process, which avoids the generation of HOCl, for example as described in WO 2002092586 and U.S. Pat. No. 6,288,248 involves the direct epoxidation of allyl chloride using titanium silicalite catalysis with hydrogen peroxide. Despite the advantage of reducing the generation of HOCl, allyl chloride is still an intermediate. The disadvantage of using allyl chloride is two-fold: (1) The free radical chlorination of propylene to allyl chloride is not very selective and a sizable fraction (>15 mole %) of 1,2-dichloropropane is produced. (2) Propylene is a hydrocarbon feedstock and long-term, global forecast of propylene price continues to escalate. A new, economically viable process for the production of epichlorohydrin which avoids the complications of controlled, chlorine-based oxidation chemistry and RCl generation is desirable. There is a need in the industry for a process for the generation of epichlorohydrin which involves a non-hydrocarbon, renewable feedstock.
Glycerin is considered to be a low-cost, renewable feedstock which is a co-product of the biodiesel process for making fuel additives. It is known that other renewable feedstocks such as fructose, glucose and sorbitol can be hydrogenolized to produce mixtures of vicinal diols and triols, such as glycerin, ethylene glycol, 1,2-propylene glycol, 1,3-propylene glycol and the like.
With abundant and low cost glycerin or mixed glycols, an economically attractive process for glycerin or mixed glycol hydrochlorination would be desirable. It would be advantageous if such a process were highly chemoselective to the formation of vicinal chlorohydrins, without production of RCls.
A process is known for the conversion of glycerol (also referred to herein as “glycerin”) to mixtures of dichloropropanols (also referred to herein as “dichlorohydrins” or DCH), compounds I and II, as shown in Scheme 1 below. The reaction is carried out in the presence of anhydrous HCl and an acetic acid (HOAc) catalyst with water removal. Both compounds I and II can then be converted to epichlorohydrin via treatment with caustic.

Various processes using the above chemistry in Scheme 1 have been reported in the prior art. For example, epichlorohydrin can be prepared by reacting a dichloropropanol such as 2,3-dichloropropan-1-ol or 1,3-dichloropropan-2-ol with base. Dichloropropanol, in turn, can be prepared at atmospheric pressure from glycerol, anhydrous hydrochloric acid, and an acid catalyst. A large excess of hydrogen chloride (HCl) gas is recommended to promote the azeotropic removal of water that is formed during the course of the reaction.
For example, Gibson, G. P., Chemistry and Industry 1931, 20, 949-975; and Conant et al., Organic Synthesis CV 1, 292-294, and Organic Synthesis CV 1, 295-297; have reported distilled yields of dichlorohydrins in excess of 70% for dichlorohydrins, compounds I and II in Scheme 1 above, by purging a large excess of anhydrous HCl (up to 7 equivalents) through a stirred solution of glycerol and an organic acid catalyst. The processes described in the above references require the use of atmospheric pressures of HCl which is used as an azeotroping agent to remove the accumulated water. Other azeotropes are known. For example, U.S. Pat. No. 2,144,612 describes using n-butyl ether along with excess hydrogen chloride (HCl) gas to promote the reactive distillation and removal of water.
Indeed, all of the prior art teaches the vaporization of azeotropes with water to provide high conversion and a process need for sub-atmospheric or atmospheric pressure conditions to accomplish water removal. U.S. Pat. No. 2,144,612 argues the advantageous use of an added azeotroping agent (for example, n-butyl ether) to promote the reactive azeotropic distillation and elimination of water, again using excess HCl at atmospheric conditions. A similar approach using vacuum removal of water is taught in German Patent No. 1075103.
German Patent No. 197308 teaches a process for preparing a chlorohydrin by the catalytic hydrochlorination of glycerine by means of anhydrous hydrogen chloride. This reference teaches a batch process with separation of water at atmospheric conditions. German Patent No. 197308 does not teach carrying out the hydrochlorination reaction process at elevated pressures.
All known prior art for the production of chlorohydrin reports hydrochlorination processes where water is removed as a co-product from the process. In particular, WO 2005/021476 teaches a series of hydrochlorination reactions in which the water of reaction is removed in an atmospheric or sub-atmospheric process by reactive distillation. Similar art is taught in WO2005/054167 with the additional teaching that the reaction carried out under higher total pressures (HCl partial pressure not specified) may improve the rate of reaction. However, nothing in WO2005/054167 discloses the use of HCl partial pressure and its effect in its process. WO2005/054167 also exemplifies the need to remove water to effect high conversion and selectivity under atmospheric or subatmospheric pressures. Neither WO 2005/021476 nor WO2005/054167 teaches any advantage of leaving water in their processes, or that removing the water effects the formation of unwanted chloroethers and RCl's. RCl's are halogenated hydrocarbon by-products of either a volatile or high boiling nature.
The use of extremely large excess amounts of hydrogen chloride (HCl) gas is economically problematic and the inherent contamination with water of the unreacted hydrogen chloride and RCl's results in an aqueous hydrogen chloride stream that is not easily recyclable. Furthermore, reaction times of 24 to 48 hours are required to achieve a far from complete conversion of glycerin; however, the products often include significant amounts of the undesired overchlorinated trichloropropane and chlorinated ethers. Other processes are also known that use reagents that convert alcohols to chlorides but that scavenge water in situ. For example, thionyl chloride can be used to convert glycerin to a chlorohydrin, as described in Carre, Mauclere C. R. Hebd. Seances Acad. Sci. 1930, 192 and may be selective, but produces stoichiometric amounts of SO2. The cost and expense of this reagent is not acceptable for the industrial production of epichlorohydrin or any other chlorohydrin derived from a multihydroxylated-aliphatic hydrocarbon. Likewise, other hydrochlorination reagents which are mild and effective are considered expensive and exotic for this transformation, as described in Gomez, et al. Tetrahedron Letters 2000, 41, 6049-6052. Other low temperature processes convert the alcohol to a better leaving group (for example, mesylate) and provide a soluble form of chloride via an ionic liquid used in molar excess, as described in Leadbeater, et al. Tetrahedron 2003, 59, 2253-58. Again, the need for anhydrous conditions, stoichiometric reagents and an expensive form of chloride prevents industrial consideration of the above process. Furthermore, these reagents can cause exhaustive chlorination of a multihydroxylated-aliphatic hydrocarbon, leading again to undesirable RCl by-products, as taught in Viswanathan, et al. Current Science, 1978, 21, 802-803.
To summarize, there are at least five major disadvantages to all of the above known approaches for preparing a chlorohydrin from glycerin or any other vicinal-diol, triol or multihydroxylated-aliphatic hydrocarbon: (1) Atmospheric pressure processes for the hydrochlorination of glycerin or any diol require a large excess of HCl, oftentimes 7-10 fold molar excess. In an atmospheric pressure process the excess anhydrous HCl is then contaminated with water. (2) Variants of the above known processes are very slow, batch type reactions, which often take between 24-48 hours at temperatures in excess of 100° C. and do not exceed 80-90% conversion to desired chlorohydrin product(s). (3) Exotic hydrochlorination reagents may drive the reaction by scavenging water, but oftentimes produce a by-product inconsistent with the economic production of a commodity. (4) All of the above approaches produce higher levels of unwanted RCls, as defined above for glycerin hydrochlorination. (5) When the reaction is run at elevated pressure to control evaporization of the reactor contents, low partial pressures of HCl result in low conversions or retarded reaction rates.
The prior art concludes that water removal is required to promote complete conversion of glycerin to dichlorohydrins. To accommodate this water removal requirement, the prior art reactions are conducted under azeotropic or reactive distillation or extraction conditions which requires a co-solvent or chaser and considerable capital addition to the process. All prior art has concluded that there is an equilibrium limitation to this conversion due to the presence of water in the reaction mixture.
It is desired in the industry to provide a hydrochlorination process for the production of high purity chlorohydrins from multihydroxylated-aliphatic hydrocarbons which overcome all of the inadequacies of the prior art. It would, therefore, be an advance in the art of chlorohydrin chemistry to discover a simple and cost-effective method of transforming diols and triols to chlorohydrins.